Loading ...
Sorry, an error occurred while loading the content.

IV-17 SS kinda working!

Expand Messages
  • shklaw75
    Hi Group, I have put together four IV-17 tubes on a breadboard with a HV5812 and a PIC18F25K22. It is powered by one of Konstantin s (kosbo) VFD power supplies
    Message 1 of 9 , May 15 1:29 AM
    • 0 Attachment
      Hi Group,

      I have put together four IV-17 tubes on a breadboard with a HV5812 and a PIC18F25K22. It is powered by one of Konstantin's (kosbo) VFD power supplies which has a variable 0-9V output for the filament (set to 2.4V), a variable 15-50V output for the anode/grid (set to 40V) and a 5V supply for the uP and HV5812.

      I am sending commands from a PC program called realterm via a USB to TTL convertor.

      The problem is that i can only get the version of the software to display!
      This is what happens. I send the command V to display the software version, v111 comes up, i then send command (MD,1,Stop) to display a word and the tubes go blank. I then send the V command again and the version comes back up again. I flashed the uP with the rev13 hex file and get the same results, v111 as the software version number. The new test function in v13 works, if take pin 14 high, all segments come on, low, all off, so it has definitely updated the flash in the uP.

      Some other commands work, i can change the font and then when i send V, the version number is shown in the new font. I can also change the effect and see that working when i go from a blank display to showing the software version or vice versa.

      I attached a 32,768 Hz crystal to pins 11 and 12 with 22pf caps and tried the clock commands, unfortunately the time does not come up.

      Not sure if this is related to what is happening above, but when i load the hex into the PICkit2 programmer, a warning message comes up saying" Some configuration words not in hex file. Ensure default values above right are acceptable."

      I have uploaded screenshots of PICkit2 screen and the configuration word editor screen to my flickr page. There is also a shot of the tubes showing the software version.
      http://www.flickr.com/photos/mrlaw/sets/72157625204679347/detail/

      On a side note, i can only get part of the PWM output working. The pot varying 0-5v on pin 3 varies the duty cycle, but the pot varying 0-5v on pin 2 does nothing. The output PWM stays at ~52kHz.

      Is anyone able to shed some light on the issues i am having above?

      Many thanks.
      Simon
    • fixitsan2
      Hello Simon, I think I have made a mistake in documenting details which are no longer relevant. Instead of sending MD,1,STOP, please try sending the command
      Message 2 of 9 , May 15 7:51 AM
      • 0 Attachment
        Hello Simon,
        I think I have made a mistake in documenting details which are no longer relevant.

        Instead of sending MD,1,STOP, please try sending the command without the commas, IE 'MD1STOP' . Note that the 1 has to be in decimal or hex format, or in other words not an ascii character 1. (Ascii '1' is decimal 49, hex $31). Instead send decimal 1 or hex $01.

        This gets confusing if you're using a PC terminal program which doesn't allow you to mix numeric types.

        MD1STOP in hex is - 4D 44 01 53 54 4F 50

        Please let me know if this has helped.
        Chris

        PS, I'll look into the PWM controls as soon as I'm back at my usual PC. One detail I have ommitted is the inclusion of some 4k7 resistors between the wiper of the pots and the AD input pins of the PIC. The value doesn't seem too critical and anything from 2k2 to 10k will work, depending on the value of the pot. I recently used 4k7 linear pots with a 4k7 resistor in series with the wiper.
      • shklaw75
        Hi Chris, That got the message display working. Either sending the whole command in hex (more of a pain to convert when typing on a keyboard!) or a combination
        Message 3 of 9 , May 16 3:28 PM
        • 0 Attachment
          Hi Chris,

          That got the message display working. Either sending the whole command in hex (more of a pain to convert when typing on a keyboard!) or a combination of ascii and hex (or dec) worked as well.

          The scrolling MS command sort of worked, do the numbers have be decimal as well? What do the numbers do? Eg. in the command MS5062STOP what does 50, 6 and 2 do?

          Is there any other command that need numbers to be send in dec/hex instead of ascii?

          I tried adding a 10k resistor in series with the pot (10k) wiper going to the pic pin, the frequency does not change on the output. Changing the pot for the duty cycle still works.

          Cheers.
          Simon

          --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@...> wrote:
          >
          >
          > Hello Simon,
          > I think I have made a mistake in documenting details which are no longer relevant.
          >
          > Instead of sending MD,1,STOP, please try sending the command without the commas, IE 'MD1STOP' . Note that the 1 has to be in decimal or hex format, or in other words not an ascii character 1. (Ascii '1' is decimal 49, hex $31). Instead send decimal 1 or hex $01.
          >
          > This gets confusing if you're using a PC terminal program which doesn't allow you to mix numeric types.
          >
          > MD1STOP in hex is - 4D 44 01 53 54 4F 50
          >
          > Please let me know if this has helped.
          > Chris
          >
          > PS, I'll look into the PWM controls as soon as I'm back at my usual PC. One detail I have ommitted is the inclusion of some 4k7 resistors between the wiper of the pots and the AD input pins of the PIC. The value doesn't seem too critical and anything from 2k2 to 10k will work, depending on the value of the pot. I recently used 4k7 linear pots with a 4k7 resistor in series with the wiper.
          >
        • fixitsan2
          Thanks for the update Simon. Sending those numbers in deciaml is the only way to give control of large displays. for upto 10 tubes then we can use ascii
          Message 4 of 9 , May 17 4:55 AM
          • 0 Attachment
            Thanks for the update Simon.

            Sending those numbers in deciaml is the only way to give control of large displays. for upto 10 tubes then we can use ascii characters 0-9, but after that, to send '11' for example, requires another byte of data. The alternatives are to either use decimal , allowing up to 255 tubes to be addressed, or use comma delimiting again. Sending decimal/hex values takes the least amount of processing time which seemed impoortant as that processing is done during an interrupt and display aberations were always going to be a risk if processing took too long.

            I have updated the manual to cover the points yoiu have raised and have copied the relevant scroll command instructions from it here.

            I cannot say why the pwm frequency is not changing because I haven't been back to my breadboard yet, but from what I recall my Fluke77 was showing a good range of frequency when I turned the relevant pot. I'll look into it, but it might be worth checking that when yoiu change the frequency, in order to get an accurate reading the duty cycle should be about 50% to start with. If you are using logarithmic pots 50% duty cycle is quite a way off to one end of travel.

            Anyway, here's the scrolling instruction. please let me know how yoiu get on with, thanks.

            <><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>

            To display the word `Stop' by scrolling it from right to left send the following command
            MS5062Stop. The sequence of numbers are custom values which are relevant to the scrolling command..

            There are three values immediately following the MS command to be aware of
            MS<1-255><start position><end position>

            The first value is the time step delay value, 1-255. A value of 255 equates to a delay of about 2 seconds between scroll steps, which is a very slow scroll effect. A value of 50 equates to about 0.4 seconds, which is quite a comfortable delay. This value can be thought of as the hold time after a scroll step has taken place, before the next step of the scroll routine is executed.

            <start position> and <end position> should be values of between 1 and the maximum number of tubes in your display. Imagine a display made up of twelve characters, and you would like to scroll a word across only the middle 4 character positions, leaving the rest of the display intact. The middle character positions are numbered 5,6,7 and 8 If you want to scroll the word `Scrolling' across only those 4 characters, from right to left, then the command would be as follows
            MS<50><8><5>Scrolling The <> brackets are not sent in the final command, so the final instruction would look like MS5085Scrolling

            What if you want to scroll from left to right instead ? This time just change the position of the numbers 5 and 8. The software knows which direction you would like to scroll in by the order of the numbers. To scroll the same word from left to right this time, send the command MS5058Scrolling.



            Note that the number of scrolling steps equals the number of characters. In order to clear a display after a word has been scrolled onto it you need to add following or preceding spaces as appropriate. For example, in order to scroll the word `Stop' across the display and then clear the display afterwards send the following command
            MS5062Stop<4_spaces>



            --- In smartsockets@yahoogroups.com, "shklaw75" <shklaw75@...> wrote:
            >
            > Hi Chris,
            >
            > That got the message display working. Either sending the whole command in hex (more of a pain to convert when typing on a keyboard!) or a combination of ascii and hex (or dec) worked as well.
            >
            > The scrolling MS command sort of worked, do the numbers have be decimal as well? What do the numbers do? Eg. in the command MS5062STOP what does 50, 6 and 2 do?
          • shklaw75
            Hi Chris, Thanks for the explanation on the reasons behind the use of dec/hex in the commands and the details of the message scrolling commands, working very
            Message 5 of 9 , May 18 4:26 AM
            • 0 Attachment
              Hi Chris,

              Thanks for the explanation on the reasons behind the use of dec/hex in the commands and the details of the message scrolling commands, working very nicely now.

              Tried the freq pot while the duty cycle pot was set to 50%, no change to the output freq. I am using linear pots, so the middle is in the middle!

              Another thing I would like to get working is the decimal points. Any chance of a schematic for wiring them up, or an updated schematic of the IV-17 smartsocket for the manual?

              Much appreciated.
              Simon

              --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@...> wrote:
              >
              > Thanks for the update Simon.
              >
              > Sending those numbers in deciaml is the only way to give control of large displays. for upto 10 tubes then we can use ascii characters 0-9, but after that, to send '11' for example, requires another byte of data. The alternatives are to either use decimal , allowing up to 255 tubes to be addressed, or use comma delimiting again. Sending decimal/hex values takes the least amount of processing time which seemed impoortant as that processing is done during an interrupt and display aberations were always going to be a risk if processing took too long.
              >
              > I have updated the manual to cover the points yoiu have raised and have copied the relevant scroll command instructions from it here.
              >
              > I cannot say why the pwm frequency is not changing because I haven't been back to my breadboard yet, but from what I recall my Fluke77 was showing a good range of frequency when I turned the relevant pot. I'll look into it, but it might be worth checking that when yoiu change the frequency, in order to get an accurate reading the duty cycle should be about 50% to start with. If you are using logarithmic pots 50% duty cycle is quite a way off to one end of travel.
              >
              > Anyway, here's the scrolling instruction. please let me know how yoiu get on with, thanks.
              >
            • fixitsan2
              Thats some progress, Simon. I m pleased you ve got scrolling to work. I ve just downloaded the same software from the files section to my breadboard and both
              Message 6 of 9 , May 18 8:34 AM
              • 0 Attachment
                Thats some progress, Simon. I'm pleased you've got scrolling to work.

                I've just downloaded the same software from the files section to my breadboard and both PWM control pots work as they should. Have you tried measuring the voltage from the pots at the pic inputs to make sure it is varying as the pots turn. Maybe the frequency pot has lost it's ground connection ?

                The decimal points are driven from pins 25 and 26. These are active high, so you need a suitable buffer or a pair of transistors typical of those used to drive nixie tube anodes, an MPSA42/MPSA92 pair.
                I don't have the time to draw a schematic now, so referencing the schematic on the last page of this manual....
                http://www.ledsales.com.au/pdf/frank2_clock_kit.pdf

                ...zoom in to the top left corner and build Q9,Q3,R7,R13 for each of the two decimal point lines. Drive the bases of the MPSA42's (Q9) form the pic via a series 22k resistor (10k-33k should work). Change R13 to 100k and try that. Instead of driving a nixie anode you will instead connect the line to the decimal points of one group , right or left. Probably R22 should be much lower, like 1k. If it doesn't work first time try raising the 10k value of R7 until it works. I used to use the transistor pair, before I used the transistors for something else which is why the values are sketchy.

                As for the clock function, that function got backbenched and needs some work still. It might work if you use decimal numbers where yoiu see numbers in the command structure but I can't say for sure what will happen, if anything.

                Chris

                --- In smartsockets@yahoogroups.com, "shklaw75" <shklaw75@...> wrote:
                >
                >
                >
                > Hi Chris,
                >
                > Thanks for the explanation on the reasons behind the use of dec/hex in the commands and the details of the message scrolling commands, working very nicely now.
                >
                > Tried the freq pot while the duty cycle pot was set to 50%, no change to the output freq. I am using linear pots, so the middle is in the middle!
                >
                > Another thing I would like to get working is the decimal points. Any chance of a schematic for wiring them up, or an updated schematic of the IV-17 smartsocket for the manual?
                >
                > Much appreciated.
              • fixitsan2
                Simon, apologies, I changed the link and forgot to change the reference to it (working in too much of a rush...) The area of the diagram concerned is top
                Message 7 of 9 , May 19 12:20 AM
                • 0 Attachment
                  Simon, apologies, I changed the link and forgot to change the reference to it (working in too much of a rush...)

                  The area of the diagram concerned is top right.

                  --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@...> wrote:
                  referencing the schematic on the last page of this manual....
                  > http://www.ledsales.com.au/pdf/frank2_clock_kit.pdf
                • shklaw75
                  Hi Chris, I understood which part of the schematic you were referring to, thanks for the clarification. The decimal points are working, although slightly
                  Message 8 of 9 , May 23 4:28 AM
                  • 0 Attachment
                    Hi Chris,

                    I understood which part of the schematic you were referring to, thanks for the clarification.

                    The decimal points are working, although slightly duller compared to the other segments. The values of the resistors I am using are 33k for the bases of the MPS42's, 100k from the collector of the MPS42 to the base of the MPS92. 10k's between the base of the MPS92 and the 40V supply and 1k between the collector of the MPS92 and the left or right group of decimal points. I also tried without the 1k and the brightness seems the same.

                    Now, I have gone and done something stupid and think I have damaged the PIC chip and maybe the HV5812! I was sticking 12V into pin 14 of the PIC (to illuminate all the segments) by accident and now i cant get anything to come up except for the decimal points! So i will get another PIC in and get it working again. I will then update you on where it is all at.

                    Cheers.
                    Simon



                    --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@...> wrote:
                    >
                    > Simon, apologies, I changed the link and forgot to change the reference to it (working in too much of a rush...)
                    >
                    > The area of the diagram concerned is top right.
                    >
                    > --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@> wrote:
                    > referencing the schematic on the last page of this manual....
                    > > http://www.ledsales.com.au/pdf/frank2_clock_kit.pdf
                    >
                  • fixitsan2
                    Such is the way of progress I suppose. We ve all been there, I m sure of that ! I m wondering if changing the values of the resitors will help you to increase
                    Message 9 of 9 , May 27 4:20 AM
                    • 0 Attachment
                      Such is the way of progress I suppose. We've all been there, I'm sure of that !

                      I'm wondering if changing the values of the resitors will help you to increase the brightness of the DP's. Perhaps raising the value of the 10k resitors and/or reducing the value of the 100k resistors. I didn't calculate the resitor values to check the biasing of the transistors is correct, merely scaled down a circuit for a higher voltage. If I get around to it I'll have another look at that.
                      I would like to get them working well because they are a very cheap optionfor driving the segments too, 22 pairs of transistors with accompanying resistors plus a pic would be all that would be required if using the generic circuit version of the software.

                      Chris


                      --- In smartsockets@yahoogroups.com, "shklaw75" <shklaw75@...> wrote:
                      >
                      >
                      >
                      >
                      >
                      > Hi Chris,
                      >
                      > I understood which part of the schematic you were referring to, thanks for the clarification.
                      >
                      > The decimal points are working, although slightly duller compared to the other segments. The values of the resistors I am using are 33k for the bases of the MPS42's, 100k from the collector of the MPS42 to the base of the MPS92. 10k's between the base of the MPS92 and the 40V supply and 1k between the collector of the MPS92 and the left or right group of decimal points. I also tried without the 1k and the brightness seems the same.
                      >
                      > Now, I have gone and done something stupid and think I have damaged the PIC chip and maybe the HV5812! I was sticking 12V into pin 14 of the PIC (to illuminate all the segments) by accident and now i cant get anything to come up except for the decimal points! So i will get another PIC in and get it working again. I will then update you on where it is all at.
                      >
                      > Cheers.
                      > Simon
                      >
                      >
                      >
                      > --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@> wrote:
                      > >
                      > > Simon, apologies, I changed the link and forgot to change the reference to it (working in too much of a rush...)
                      > >
                      > > The area of the diagram concerned is top right.
                      > >
                      > > --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@> wrote:
                      > > referencing the schematic on the last page of this manual....
                      > > > http://www.ledsales.com.au/pdf/frank2_clock_kit.pdf
                      > >
                      >
                    Your message has been successfully submitted and would be delivered to recipients shortly.